You’re likely coming to serverless cloud for the first time from one of a few different directions. You may have reached a point where your legacy systems can’t meet the business or your customer’s demands for speed and flexibility. Maybe, the time is right to migrate off of your iPaaS.
You’ve investigated serverless, and believe it has potential. But you need to prove the value. And you can’t invest months to find out if it’s the right path for your organization.
A proof of concept (POC) project can give you the evidence you need to explore the path to the serverless cloud. You can be up and running with a POC in weeks, and the process of getting there will uncover your company’s strengths and challenges in addressing a full migration.
Preparing for Your First Serverless Cloud Integration Project
Approaching cloud migration by attacking a POC first offers a variety of advantages. First, you can use the proof of concept alongside your existing integration to try it out before fully committing. A POC uncovers issues and pain points while the stakes are lower.
Failing fast with a POC also sets later iterations up for success and lets you develop future plans that avoid exposed pitfalls. It also provides something tangible to demonstrate to the rest of the organization to prove the value of a cloud-based solution.
Regardless if you’re starting from an existing iPaaS or a legacy system, your process to define the right candidate for your POC will be the same. Decomposing your systems into smaller components will provide a clear picture of your candidate integrations. Below is a roadmap for planning and executing this first project.
Keep in mind, though, this first step is crucial, and shouldn’t be short-cut or skipped. Big Compass frequently helps customers at this phase effectively plan a migration strategy and break down processes, interfaces, and components of a system to identify ideal candidates to move to the cloud first.
The planning phase starts with an inventory of all of your existing processes. Using this view, rank the integrations based on risk, complexity, volume, frequency, cost, and importance. This information will let you evaluate and collect data for the best POC candidate based on these dimensions. An example of a good candidate to move to the serverless cloud first is low complexity, high volume, varied frequency processes. Another candidate could be integration infrastructure costing you the most due to over-provisioning or scaling events.
Process to Migrate
Typically, you’ll want to choose a process for your POC that’s easy to migrate. It will give you the ability to learn as you go and better explore the tools and technologies you’ll need while at the same time reducing the risk of the project. However, if there is a burning need that elevates another process – a critical deadline, a customer issue, etc. – you can certainly choose the highest priority process, instead.
Time and Money
If you’re migrating from an iPaaS provider to reduce costs, that renewal deadline should drive your project. Also, take the time to consider the ROI of your project. How aggressive do you need to be to reach a break-even point? ROI can be a critical driver for how aggressively you want to migrate to the cloud. The quicker the migration effort, the higher the upfront capex spend on the migration itself, but the sooner you realize an ROI. You may choose an aggressive migration timeline for this reason, a relaxed migration timeline for an extended ROI due to immediate budget concerns, or you may land somewhere in the middle to balance capex and opex spend.
Customer Communication and Interaction
Inventory customer contact information along with your process matrix. This will simplify communication for externally facing endpoints. One of the primary pain points for our clients going through a migration is tracking the migration status. Creating a well-defined inventory will help you stay organized as your migration effort ramps up.
Staffing and Expertise
Having the proper domain expertise – AWS, Azure, etc. – will also determine your POC choice, as it impacts the decision around almost every other dimension. You have three options to leverage domain expertise to get your POC off the ground:
- Internal experts
- Hire experts
- Bring in a consultancy like Big Compass
Using internal resources comes at the cost of other projects. Hiring an expert commits you to a specific path. Consultants, of course, also come with a cost, but they give you the flexibility to evaluate implementation and migration choices without losing ground on existing IT projects.
The design phase for your POC will mirror many other IT projects. Once you’ve chosen the right candidate process to migrate, you’ll need to develop its requirements. Business stakeholders can offer up business requirements, while legacy or iPaaS SMEs can cover the technical requirements.
Now is the time to define your software development lifecycle (SDLC) protocol as well. Take the time to think about seamless integrations, deployments, preventing errors, and what your CI/CD pipeline should look like. This step is going to be technology agnostic – it will make your teams successful no matter what tech you decide to use.
The design phase is also time to make the all-important choice of how you’ll get the POC done. You have a few options to choose from here, as well.
Using knowledge from internal teams can get you started. They’ll understand the business, and possibly even the specific process being migrated. Internal resources may only get you so far, though.
At some point, you’ll need expertise that may not exist within your current skills warehouse. If that’s the case, a blended team might get you to a functional solution faster.
Blended teams put your internal resources side-by-side with external experts. The internal teams can architect data flows while the external resources can train and coach staff during the implementation.
Of course, if time is of the essence, an external team can rapidly get you to a working POC. Professional services, like Big Compass, hit the ground running and can move you quickly to and through the implementation phase.
While design is a critical component to any project, you can choose to cut out certain phases of design to speed up the implementation and get you to the cloud quicker. If a POC in the serverless is the highest priority and requirements for the integration are well known and documented, a technical team can architect the cloud solution and begin the implementation quickly. This could allow you to prove the solution and fail fast, while in parallel (or after the initial implementation) you define the full SDLC, CI/CD, development standards, and all of the other processes that will set you up for long-term success. Balance your short-term goals and iterative success with long-term success when in the design phase.
Like the planning phase, you have three options to get the code written – internal teams, a blended resourcing model, or using professional services to get the work done.
What’s crucial at this point is having made the first two phases airtight. With solid planning and design, you can easily find people to implement your POC. Without that legwork upfront, your only real choice will be your internal team.
Since you’re porting an existing process, not much should change once you get to testing. You should have similar hooks as you did before for QA testing, regression tests should replicate the legacy system.
This is an excellent time to establish your testing protocol and framework for services on the new infrastructure, however.
For the initial POC, deployments, and monitoring can happen manually for now. You will want to consider automated vs. manual deployments for the future and keep tabs on what you’ve learned from your proof of concept project about both. Refer back to the decisions you made around your SDLC back in the design phase.
Management of your deployments will be on-going. One of the benefits of the cloud is the comprehensive management features that are available. The deployment phase should include considerations for using the security and monitoring features your cloud provider offers.
Security should be examined from multiple angles. Things like network security are primarily handled by the provider, while access and role management can be accomplished with IAM tools.
Functionally, monitoring can be set up for automated alerting and even ticket creation. Alerts can also be established around cost thresholds to keep budgets in check. Creating a dashboard based on key indicators makes it easier to understand system health, throughput, and costs.
Using your first POC to test out and explore your new serverless environment will let you imagine its potential as a tool to meet the needs of both IT and the business. The POC also provides a chance to get teams up to speed, build repeatable processes, and create foundational assets for continued migration to the cloud.
If you’re looking for help to get your serverless cloud migration moving in the right direction, Big Compass has multiple ways to support you. From coaching all the way to full-on design and implementation, we can help your organization quickly realize the benefits of the serverless cloud.